[13:24] <balloons> poor testdrive stuck in a gtk2 world
[13:29] <flocculant> could be worse, could be tesdrivealada ...
[13:29] <DanChapman> :-)
[13:30] <flocculant> hi DanChapman :)
[13:30] <DanChapman> hey flocculant
[13:40] <nuclearbob> DanChapman: I'm getting an error when trying to run the lvm test: http://pastebin.ubuntu.com/12408554/ I'm taking a look, but if you have any ideas, let me know
[13:44] <nuclearbob> DanChapman: I may be invoking the test case incorrectly. I'm looking at that further
[13:45] <DanChapman> nuclearbob: yeah it looks like it maybe being invoked from the wrong directory?? the module import is correct though "ubiquity_autopilot_tests.tests.test_nonenglish_lvm" is correct
[13:46] <DanChapman> nuclearbob: make sure your in /ubiquity/autopilot
[13:47] <nuclearbob> DanChapman: I'm in there. My command line is: ./ubiquity-autopilot-runner/run-ubiquity-test -T testrunner.cfg --sdl -s ~/iso/ubuntu/wily-desktop-amd64.iso  -t unittest.loader.ModuleImportFailure.ubiquity_autopilot_tests.tests.test_english_lvm
[13:47] <nuclearbob> er
[13:47] <nuclearbob> hey, I can see the problem with that right now
[13:47] <nuclearbob> I just used what autopilot.list gave me
[13:47] <DanChapman> :-D yeah you need to drop the unittest.loader.ModuleImportFailure.
[13:47] <nuclearbob> and I'll make sure autopilot is updated and see if I need to file a bug there
[13:48] <flocculant> nuclearbob: you had any luck with a flavour image at all?
[13:52] <nuclearbob> flocculant: that's next after I get the main changes landed
[13:52] <nuclearbob> but I suppose I may as well start downloading some so they'll be ready
[13:54] <flocculant> xubuntu is obviously the easiest one to get :D
[14:03] <nuclearbob> flocculant: oh good, that's the one I started getting :)
[14:03] <nuclearbob> DanChapman: problem resolved if I invoke things correctly
[14:03] <flocculant> nuclearbob: \o/
[14:03] <flocculant> ha ha
[14:04] <flocculant> iirc it's only studio which is silly size
[14:37] <nuclearbob> flocculant: the test I ran on xubuntu looks good
[14:37] <flocculant> \o/
[14:38] <flocculant> though I assume it failed - there's an issue currently with images
[15:07] <sak> So, I am setting up a new GPG key to be registered on launchpad. This key will be used to sign the code-of-conduct. I did have a key registered but when trying to sign the file, I was getting a "secret key not found" error in the terminal. Reading through the opengpg key instructions, my system could not find any keys registered, local or on a server. I started fresh. Do I have to wait 30mins for my new gpg key to be register
[15:07] <sak> ed on launchpad?
[15:10] <davmor2> sak: possibly
[15:13] <sak> Ok. The waiting game begins. I will wait until this key is on file before I use bzr
[15:31] <DanChapman> nuclearbob: \o/ awesome! I still need to test edubuntu that has a couple of additional pages during the install, and it has a special partition config for the custom_install tests so need to check that works as well
[15:33] <balloons> sak, ahh yes, changing keys is fun!
[15:34] <flocculant> that'll be some really bizarre version of fun :p
[15:34] <sak> lol balloons,
[15:34] <balloons> I went through the same thing with ssh keys
[15:35] <balloons> and then transfering them securely between machines, or making new ones, etc
[15:35] <balloons> new installs wipe the old keys
[15:35] <sak> fun will be getting everything working and breaking it again right ballons?
[15:35] <balloons> I enjoy making things work again instead of breaking :-)
[16:01] <sak> How long does it take for launchpad to send an email with instruction to complete the gpg key registration?
[16:02] <flocculant> depends on whether people fed the hamster running the power generating wheel
[16:02] <flocculant> usually quite quick iirc - check spam etc
[16:06] <sak> Nothing yet
[16:26] <balloons> sak, still waiting?
[16:27] <balloons> what's your lp id?
[16:27] <sak> saqman2060
[16:37] <balloons> sak, what does https://launchpad.net/~/+editpgpkeys show for you?
[16:43] <balloons> I ask, because I see no keys for you. LP should email you pretty quick about signing and returning the mail. You did that?
[16:43] <balloons> If so, I would expect to see a key there
[16:45] <sak> checking now
[16:46] <sak> yeah, the only key that I see there is the deactivated one
[17:16] <dkessel> balloons: what's the status on the jenkins server for the auto image tests? who was going to set the server up?
[17:30] <flocculant> dkessel: max apparently https://lists.ubuntu.com/archives/ubuntu-quality/2015-August/006106.html
[17:34] <balloons> dkessel, yep, that's correct. Max, aka nuclearbob
[17:35] <balloons> if you've been seeing the conversation in IRC, he's close to replicating things looks like. I've anxiously awaiting success
[17:36] <dkessel> ah, great! no, i haven't been able to follow really due to connectivity and still lots of stuff to do after moving
[17:37] <sak> Ok, I am heading to work. Will continue this later
[17:47] <nuclearbob> balloons, dkessel, flocculant: I
[17:47] <nuclearbob> 've gotten some good manual runs with flavors today. I'm working to get DanChapman's branches landed and looking at the automated jenkins job creation in the branch to try to get that running here
[17:52] <flocculant> nuclearbob: that's awesome news :)
[18:17] <flocculant> balloons: not sure if I'm reading this right ... http://ci.ubuntu.com/smokeng/wily/desktop/amd64/20150914/13881/
[18:17] <flocculant> if that includes booting the image - why is it not showing fail?
[18:18] <balloons> which one?
[18:18] <flocculant> reason for the comment is that currently images fail to get to desktop
[18:18] <balloons> those are all preseeds
[18:18] <flocculant> oh ok :)
[18:18] <balloons> :-)
[18:18] <flocculant> basically nuclearbob said earlier he was getting things to work - but currently images are fubar
[18:19] <flocculant> so ...
[18:19] <flocculant> wondering if the test is actually what we want
[18:20] <flocculant> <nuclearbob> flocculant: the test I ran on xubuntu looks good
[18:20] <flocculant> but boot the current daily and apport starts :)
[18:47] <nuclearbob> flocculant: hmm. My xubuntu test should have been on the current daily. We
[18:47] <nuclearbob> 're also looking at getting post-install smoke testing running regularly again
[18:49] <flocculant> that's good news :)
[18:50] <flocculant> nuclearbob: it's not just X that doesn't boot without apport whining - Ubuntu is the same
[18:50] <flocculant> so if you're getting green lights - then something is awry somewhere :(
[18:51] <nuclearbob> flocculant: I'll make sure I've got the latest ubuntu image and run again
[18:53] <flocculant> nuclearbob: as long as it's an image from the last day or so - it should fail
[18:55] <nuclearbob> flocculant: it should be up to date, but it could have been a few days old. I'm running with the new one now
[18:59] <nuclearbob> flocculant: I don't see any errors when I run a ubiquity test on today's ubuntu image with md5sum 93238d6e2efe15f662dd86538b913c00
[19:04] <balloons> flocculant, what's not working for you on today's image?
[19:05] <flocculant> nuclearbob: same image
[19:05] <flocculant> balloons: 2 ticks
[19:05] <flocculant> https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1495017
[19:05] <flocculant> that's what they all show
[19:06] <flocculant> or at least - xubuntu/gnome/lubuntu and ubuntu do
[19:06] <flocculant> not looked at any others
[19:06] <flocculant> my concern here is that - out in the real world there is an issue which stops them booting properly
[19:06] <nuclearbob> flocculant: I guess I'm not using the ubiquity trunk, that might be a difference. I can try that, but the tests will fail on that at present
[19:06] <balloons> I believe the difference here is that we go in via the live session
[19:07] <balloons> which works according to the bug
[19:07] <flocculant> balloons: they do - once you've ignored apport and then clicked continue
[19:07] <flocculant> all I'm getting it is that out in the world images fail
[19:07] <flocculant> the autotest passes
[19:09] <flocculant> nuclearbob: not sure what's going on - but it might be worth keeping an image known to fail around just to prove later
[19:10] <nuclearbob> flocculant: yeah, that seems like a good idea if you've got one
[19:10] <flocculant> I can keep some here for sure
[19:11] <flocculant> balloons: mmm - I see what you mean :|
[19:11] <flocculant> this is an issue
[19:11] <flocculant> I can get to desktop live
[19:11] <flocculant> hit install and it fails
[19:11] <flocculant> so it STILL fails - just a bit later :)
[19:11] <balloons> we don't invoke the desktop launcher
[19:11] <balloons> which might come into play here
[19:11] <flocculant> yea
[19:12] <knome> though aiui, we don't even launch any desktop with xubuntu any more
[19:12] <knome> we just run feh to draw the wallpaper
[19:12] <balloons> we can't complete replicate things, hence the need for humans :-)
[19:12] <balloons> but yes, I like the idea of trying to replicate the breakage
[19:12] <knome> if that change has landed already, that is
[19:12] <flocculant> balloons: the thing is though - rarely do people hit a key when the image boots
[19:12] <flocculant> they get to the try/install thing and choose
[19:13] <flocculant> and it is failing there
[19:13] <flocculant> balloons: I understand that - but should we not at least test that install is possible?
[19:13] <flocculant> knome: not yet we don't
[19:14] <flocculant> pretty sure we're waiting for ubiquity changes
[19:14] <knome> aha
[19:14] <knome> ok
[19:14] <knome> for some reason i got the impression that this was done... but ok :)
[19:14] <knome> maybe that changes things again, who knows
[19:17] <flocculant> balloons: so what exactly does this test test currently - just that the image starts?
[19:18]  * DanChapman reads scrollback to catch up
[19:19] <flocculant> if only I'd taken the chance to get coding training and a job for free in 1980 - unfortunately £5k redundancy was more fun at 21 :D
[19:20] <knome> lol
[19:20] <flocculant> DanChapman: there's not much actually - can tell you in 2 lines :)
[19:21] <DanChapman> flocculant: a bit old but here's a video of one of the tests running
[19:21] <DanChapman> https://vimeo.com/95295326
[19:21] <flocculant> DanChapman: ok
[19:21] <flocculant> so that *should* fail today
[19:23] <flocculant> only need to see the start :p
[19:41] <flocculant> now I'm even more confused as to what these tests do - if they do as the vid suggests then they should definitely be failing
[19:49]  * balloons catches up again
[19:51] <balloons> flocculant, hmmm. So it should crash on start yes?
[19:51] <flocculant> yep
[19:52] <balloons> so yes, the preseed tests wouldn't show this. The AP tests I agree should. However, as I said we don't invoke things the exact same way as a user does
[19:52] <balloons> I'm guessing since it's a permissions thing, and the listed workaround using sudo is present, that's why we aren't seeing it
[19:52] <balloons> if i remember correctly, we are running as root
[19:53] <DanChapman> Yeah it runs as root, it basically does "sudo dbus_launch ubiquity --autopilot" and the test picks up the exported session bus address to hook on to
[19:54] <flocculant> mmm
[19:54] <flocculant> trying to see a positive here - help me out :)
[19:56] <balloons> flocculant, this specific bug wouldn't be caught by the tests. But in general I think most bugs will
[19:56] <flocculant> that's helping :)
[19:56] <balloons> this specific bug wasn't caught because it is a permissions thing -- when you are root, you get to own everything.
[19:57] <flocculant> I understand that we're never going to catch everything with these tests
[19:57] <flocculant> yep
[19:58] <balloons> right. But I still think it's valid to look at each bug we don't catch and consider ways to include it
[19:58] <flocculant> ok - that's positive
[20:00] <flocculant> balloons: basically I knew there was an issue, so when I was seeing 'this is all good' thought it best to say something now
[20:00] <flocculant> rather than in a week - oh yea - I saw that ...
[20:03] <balloons> definitely
[20:03] <balloons> something to bring up each time a discrepancy occurs
[20:07] <flocculant> yep
[20:13] <flocculant> I'll leave you in peace for a while now :)
[22:46] <knome> balloons, hoi?